The early stages of sensory processing pose similar problems for various sensory modalities-for example, controlling gain and minimizing noise. It is not clear how the computations that solve these problems are implemented at the level of cells and synapses. The early olfactory system is a useful preparation for investigating these issues because of its compartmental architecture. All the olfactory receptor neurons (ORNs) that express the same odorant receptor converge on the same compartment (glomerulus), and glomeruli are linked by both inhibitory and excitatory lateral connections. This architecture raises specific questions about the in vivo function of synaptic interactions in this circuit. Specifically, why do so many ORNs converge on each glomerulus? How do postsynaptic neurons integrate converging ORN spikes in the time domain? What happens when the connections between glomeruli are abolished? Do different glomerular processing channels perform different computations on their feedforward inputs? Why is there such diversity among local neurons (LNs)? Why would it be useful to have both excitatory and inhibitory LNs? These questions will be addressed using targeted genetic manipulations, in vivo whole-cell recordings, and calcium imaging in the Drosophila antennal lobe. The Drosophila antennal lobe is a good model for addressing these questions because it bears a strong similarity to its vertebrate homolog, the olfactory bulb. Moreover, it enables experiments that are currently not possible in other preparations. Specifically, it is possible to genetically manipulate specific synapses, validate the cellular correlates of these perturbations using in vivo intracellular electrophysiological measurements, and examine the functional consequences of these perturbations for the intact circuit. In these studies, genetic tools will be used (1) to manipulate ORN convergence and coherence, (2) to selectively abolish lateral excitation, lateral presynaptic inhibition, and lateral postsynaptic inhibition, (3) to stimulate specific LN populations, and (4) to monitor and manipulate the spatial spread of lateral inhibition. Most of these manipulations were made possible by recent discoveries about the biology of this system. All the techniques in these studies are routinely used in the laboratory, and thus their feasibility is proven. The results of these studies should illuminate general principles underlying synaptic integration in sensory circuits in vivo. Namely, these studies should help clarify how neural circuits can maximize their signal-to-noise ratio, how different circuit modules might perform specialized computations, why local interneurons are so diverse, and why lateral excitation and lateral inhibition often co-exist. More specifically, these studies should clarify the synaptic basis of olfactory processing, in vertebrates as well as in simpler model organisms. Understanding how the brain processes odors should aid the design of so-called artificial noses, sensors designed to analyze organic volatiles which have important applications in medical diagnosis.